Going Rogue | Hacker News Reader

Going Rogue

filfre.net

173 pointsdoppp70 comments

For me the funnest way to play nethack was with a copy of the source nearby. It's cheating of course, but it is the fun sort of cheating. I always thought of it like being a monk or wizard, journeying forth armed with the hidden secrets of the world.

I never did beat the game, but My most memorable experience was once when I was messing around in wizard mode(a sort debug/creative mode to try things out in) and summoned a high level demon for grins and giggles, which promptly killed me. I did not think much of it at the time. but nethack, when you die, writes a bones file with what you had on what level and what killed you. it normally lets you find what killed you and a chance to get your stuff back. But now my nethack install is tainted with Juiblex, the faceless evil, on dungeon level level 2. I probable could have fixed it but it made the game hilarious, every play through containing a mad dash through level 2 and fervent usually unanswered prayers that the infernal thing would not follow you to the next level.

In my opinion, source diving and using spoilers aren't cheating, if only because the games are so thoroughly randomized. You can know everything about your inventory, but one misstep and you're toast. Any ascension is as much a matter of luck as it is of skill.

> Juiblex, the faceless evil, on dungeon level level 2.

LOL! Wands of polymorph can result in interesting monsters showing up in early-level bones files. Nothing like waltzing down to level 3 and getting digested by a purple worm named Spud...

> Any ascension is as much a matter of luck as it is of skill.

There are people who can consistently ascend on nearly every run, if they pick a favorable class.

It requires extremely careful play and encyclopedic knowledge of all the little tricks to avoid danger. There are a lot of them in the game, and so long as you remember that the game is turn-based and that so long as you are careful, you can almost always just leave if things look hairy.

I played Nethack a lot during school. The only obstacles to ascending any character were my free time, patience, ability to stay focused, and occassionally the unfavorable early-game RNG.

When I was a student, a lot of people played on one guy’s machine. I got lucky with a polymorph trap and by dog became a Xorn. Inevitably I died and that bones file became the absolute bane of the other player’s existence.

Someone finally beat it and then, of course, conscripted me to spend the next half hour trying to remember what exactly I’d dropped.

> Any ascension is as much a matter of luck as it is of skill.

Is it? Admittedly, I've only fooled around with NetHack occasionally, but from observing good players it was my impression that with encyclopedic knowledge of the game, ascending can be done pretty consistently, barring some _incredibly_ unlucky happenstance or attention-related carelessness.

I watched an extremely cool talk in college where a professor showed off some of the functionality of gdb by hooking it up to a running copy of Nethack and using it for various cheating. It was a bit over my head at the time but still felt pretty magical.

I was a save-scummer. I played Nethack locally, and save-scumming was the only way that I could make meaningful progress beyond the noob levels. Yeah, I knew all the spoilers by heart, I knew the secrets, I had access to the rumors and the Wiki, but I had some bad habits that led to YASD every time, such as hoarding everything in inventory, melee attacks when ranged would be better, not making much use of spells or magic, and I would often forget that I was carrying a particular item that would be perfect for a particular situation, so I'd die with it still unused.

So save-scumming was my solution. Sometimes I played in Wizard mode, but I enjoyed save-scumming immensely, and it allowed me to get quite far, especially to the Castle and drawbridge level, and even sometimes down to Gehennom.

I did reach the Astral Plane a few times, but I never actually ascended as far as I can recall, but frankly, after the Castle level, the game gets kind of boring and unfair for me. I never liked the Wizard's harrassment as you climb back up the dungeon, level by level. Also all the curses really got me down. Save-scumming can really only take you so far if you're not an expert at planning and execution too.

I quit playing Nethack entirely a few years ago. I have moral objections to worshipping false gods, even if it's "just a game". But I bow to the genius of the Developers Who Think of Everything, especially Ken Arromdee, who's played the same MUDs/MUCKs as me, and is also an occasional Wikipedia editor.

Curses can be addressed by: 1) blessing everything that matters before you start your ascension 2) carrying inventory in an oilskin sack (which will work fine while cursed, and protect items inside it from being cursed) 3) carrying holy water and/or scrolls of remove curse 4) knowing the remove curse spell 5) certain Artifacts and armor will reduce the effect of curses on your inventory 6) carrying a lot of lightweight junk as a curse "buffer" so that curses will have a lower chance of hitting something critical 7) wielding Magicbane, that blocks 95% of curses.

A lot of this is really just, as mentioned, time spent on planning and execution.

Great article, but it did give me the "an elephant is like a rope" feeling—it seems like it's one person's idiosyncratic perspective, some of which I agree with and some I disagree with, and there isn't really much justification for changing my mind about the parts where we differ.

A couple of specific things that jumped out at me, as someone who lived through this (and got hooked for varying amounts of time on several roguelikes and roguelites):

I strongly disagree that today's world has the best of open source & proprietary. (Also, omitting any mention of Free software is a little odd.) We do have a mix, but it's neither the best nor the worst to be had. The proprietary mindset has infected the "programmers with free time" community, to the extent that one reason why you don't see ambitious user-facing projects is because the profit motive has drained away enough energy and talent that the engine of open source development just kind of sputters along until things start to risk taking off, at which point someone will come up with an idea for monetization. And monetization kills the "true" open source spirit dead.

On a related note, pointing fingers at open source games for lackluster design and poor user centric thinking is a little ironic, given that the exact same thing ails today's commercial game scene. Loot boxes and more subtle forms of the same thing have largely destroyed the freedom and original thinking that used to create interestingly novel experiences. We had a period where the indie scene carried that torch, but it has now been subverted in much the same way as the open source scene.

There are plenty of individuals or small groups out there who still have the fire, but it seems like getting any traction (whether by growing a community or gaining users/players) brings those people in too close of contact with the poisonous influence of profit or fame or whatever, and it takes a strong character and quite a bit of privilege to resist those influences. Temptation is everywhere.

> The proprietary mindset has infected the "programmers with free time" community ...

I don't know whether there is a difference between today and twenty something years ago. Anyway, as long as one has not enough money to live on passive income, time is money. So one has to somehow monetize on one's one time and skill. This is not poisonous. Earning money is just a matter of life, because one needs money to live from (or someone else has to pay the bills).

I had a quite popular open source project about two decades ago. A lot of people used it (there were more than 300 people on the projects mailing list alone), and many of them also for commercial purposes. There was an option to donate to the project; but when I remember it correctly, I recieved only a single donation in all the years. However, I considered my open source work as a kind of quid pro quo for all the free software I use. But when the project ended, I had no motivation to invest any more time in another open source project, because I considered my "debt" repaid and moved on to other kinds of social engagement.

Having been involved in at least some degree over that period, yes there is a large difference. Not a simple one; there are both good and bad differences. (I would guess that the raw amount of open source code produced today is rather larger, for example.)

> So one has to somehow monetize on one's one time and skill. This is not poisonous.

It is not poisonous in general. I get paid for programming, and don't have a problem with others getting paid for programming. But the "monetize all the things" mindset is indeed poisonous to the original heart of the open source community, where people are motivated by the desire to share and are making decisions that are consciously aligned with growing the strength and health of this open source community.

They can coexist, but there is tension and cannibalization between them[1]. It's more of an ecosystem. Two populations are living side by side with blurred edges between them. Some rate of defections from one to the other. An environment that encourages one and not the other can hugely alter the size and health of one of the populations, and that's what we've seen happen: many aspects of the environment have shifted to encourage the profit-driven crowd, to the point where the share and share alike crowd are looking anemic. It's not about one being good and the other bad, it's more like an invasive species lowering biodiversity and making all of us worse off.

Think of it as value creation vs value capture, if that works better for you. My claim is that the community has swung too far in the value capture direction. Most work these days that looks like the older model is actually being funded by massive value-capturing organizations; they're just capturing the value from somewhere else. I applaud such contributions, but don't pretend that they're passion projects of civic-minded individuals anymore.

I'm not going to look down on someone who works for money. It's not a bad thing to do, and there are good reasons why they might even want to do things differently but cannot. It's just that they're not part of the particular community I care about, and I am interested in the health of that community.

[1] Ok, monetization generally cannibalizes open source and not the other way around.

The criticism of open source software the author presented is a lot more nuanced, and I for one agree with it. Even long before loot boxes etc, the overwhelming majority of "original thinking that used to create interestingly novel experiences" in gaming did not originate from the OSS community.

It is a completely different story when it comes to non-gaming software though.

     Great article, but it did give me the "an elephant is like a rope" 
     feeling—it seems like it's one person's idiosyncratic perspective, 
     some of which I agree with and some I disagree with, and there 
     isn't really much justification for changing my mind about the 
     parts where we differ.

But this is exactly Jimmy's writing style and the entire (extremely long) series is like that - it's not meant to be a wikipedia page.

Sorry, I was being too indirect.

The whole intro section is weirdly selective about what to include and weave into a single narrative. I lived through that time. There were a lot of other things going on that paint a different picture. Not that the description is definitively wrong or anything, it's just that it is very like the fable of the blind men and the elephant. It's not wrong to say an elephant is like a rope, but it is incomplete to the point of being deceptive.

The odd rant about open source makes quite a tenuous leap to the meat of the article, which I enjoyed much more and didn't have any particular issues with. "An elephant is like a rope, now let's talk about the fascinating physiology and ecology of the hippopotamus, a roughly elephant-shaped animal."

Ok, that's unfair, the opening and the Rogue history are more related than that. But note that the latter is explicitly called out as an "exception that proves the rule" of the former. Yet "most A are B, so now let's look at a rare example of a not-B that is an A" only supports the opening thesis insofar as you explain or at least make the point that `A & not(B)` is uncommon and illustrate it with your example. Simply giving an example of where your argument doesn't hold, well... it just undercuts the argument.

Anyway, it's not the author's responsibility to write an article in a way that works for me. And I certainly don't want something complete or wikipedia-like. It's just... odd. It talks about how open source failed to live up to its expectations. That relies on an overly literal acceptance of those expectations, because in my mind it has vastly exceeded those expectations. People were indeed overenthusiastic, but even then not many were claiming it would take over everything.

Is a developer scratching their own itch a fundamentally flawed approach that will never produce software that takes its users' actual needs into account? I would argue that they are reasonably independent from each other. Someone scratching an itch may very well be able to maintain a coherent design, or something may start with itch-scratching but then attract the involvement of designer-architects, or the reverse: itch-scratchers may join in a more "designed" project once it starts to show promise and be useful enough to trigger its own itching. On a related note, the purist "cathedral vs bazaar" interpretation is largely a myth. The bazaar pretty much always evolves some level of organization, and cathedrals are staffed with people, people who are often motivated by itch-scratching no matter what the high priests ordain.

I am sympathetic to the argument that programmers tend to produce things to be consumed by other programmers. I am also sympathetic to the argument that once a program gets to the point where it is useful to other people, it has a strong tendency to get absorbed into an organization that makes it usable by a broader audience. Is that a failure of open source or a success? It depends where you draw the finish line.

Blender, Godot, Firefox, GNOME, WordPress, GnuCash... it's not like there aren't counterexamples out there.

And the plain fact that nearly everything is now built off of an open source base, and that base is steadily moving up the stack, makes me suspect the "open source can never handle X" argument has a limited shelf life.

> Blender, Godot, Firefox, GNOME, WordPress, GnuCash

Two of your examples begin with GN, which, of course, is a reference to GNU (explicit in the case of GnuCash) and thus a reference to the Free Software movement. Which the author of the piece rather noticeably neglects to talk about at all.

I agree that the open source section was a bit strange and not really necessary for the article's actual focus - when I read it I immediately thought this will create a completely avoidable flame-war.

Have been really enjoying Brogue lately. Besides the difficult, turn-based gameplay (decision making is more cerebral than reactive), I think ASCII graphics are part of what makes these games so enrapturing. Our brains are incredible at constructing rich narratives from limited information.

Brogue is genuinely beautiful. I know pure roguelikes should be white ascii on black, but it's incredible what a little color can do. And the way it handles dynamic events in its environments. Really makes the cave "biome" come alive.

Even the original Rogue had color and graphical tiles as soon as it was ported to systems that supported them. I'm sure there are people out there who think "true" roguelikes have to be monochrome but I'm lucky to never have met them.

As a rogue addict, nice read. But rogue is a gateway drug to the nethacks :)

* ssh nethack@alt.org

* ssh nethack@hf

* ssh play@server.gnollhack.com

* ssh slashem@slashem.me

Newcomers to old-school roguelikes may find Dungeon Crawl Stone Soup more accessible as an entry point to the genre, which can be played in a graphical interface in your web browser via https://crawl.develz.org/play.htm or via classic SSH interface at joshua@crawl.akrasiac.org (password "joshua").

I've been playing DCSS super casually, on-and-off, for well over a decade now, and the last time I checked in, it felt like they've perhaps sanded too many of the sharp edges off. I enjoyed the hunger system. butchering creatures into chunks for my kobold to eat was fun. throwing potions was cool. the identification minigame was good classic roguelike fun, and cursed equipment was fun. I could go on and on.

I understand that they have to keep updating it to keep it fresh for long-time super-experienced players, but it's like an entirely different game now, compared to the one I first played years and years ago.

The online servers for DCSS often host old versions if you'd prefer to play those; crawl.akrasiac.org hosts versions back through 0.11 (released in 2012). Given that it's so easy to play old versions, I'm happy that the devs are open to toy with getting rid of sacred cows in the new versions.

Also, item identification still exists, even if everything else you mentioned has been changed in some way. :P

identification is still a thing, but iirc cursed items aren't a thing anymore?

but yeah, it's nice to be able to go back and play older versions, but at this point I don't even know which version I would like to play. I'm fine with playing trunk here and there, it's just interesting to see a long-term project like this gradually shift focus over time.

I just liked playing as a kobold berserker, a cannibalistic little bastard tooling around a dungeon, getting into trouble, killing stuff and chopping it into chunks to eat, and occasionally being rewarded by my blood god.

You can install it locally as well and play in your terminal.

https://crawl.develz.org/download.htm#linux

DCSS is more combat oriented, Nethack/Slashem about thinking out of the box and doing crazy stuff to survive in critical places/situations.

(hf here is hardfought.org)

Thanks, I forgot I had hf defined in ~/.ssh/config

I cannot edit my post, so I hope people see your post

This is all I need, I'm already trying to break my Larn addiction. (/s)

I liked Larn (and Omega). I wasted many hours on Larn.

Author didn't go into the fact that some of these games arrived with PCs; AFAIK there was never a Unix version of Larn.

I thought larn was unix-only

EDIT: https://en.wikipedia.org/wiki/Larn_(video_game)

(I think I recall playing larn because it was one of the roguelikes available on unix)

Back in the day, I played the Amiga port of Larn with little graphical tiles. Lots of fun. There was a graphical Amiga port of Hack too, but it was too buggy.

Whatever you do, don't visit https://larn.org. You'll never get that time back

That’s not at all true. I played Larn on an Ultrix VAX 8600 circa 1987.

My wrong. It was written for Unix.

Well, I guess there's something funny with the license; it's not packaged for Debian, and I can't remember a time when it was.

Larn has a noncommercial license, so it's nonfree by Debian standards.

This appears to be the licence for Larn:

https://github.com/primeau/Larn/blob/master/LICENSE

That seems to be a normal MIT licence (without advertising clause). Debian packages loads of software under that licence.

[Edit] That appears to be JS Larn, which is apparently derived from Larn and uLarn, but with an MIT licence replacing the original author's. You can't do that.

I think Larn had a free clone called ularn.

Ditto, I play Slashem daily. My deepest level it's been 13 or 17, but I won't surrender. Also, IF games and csokoban games.

Since it’s a hard google: Interactive Fiction

Why not just connect to a MUD?

That's a completely different kind of game. Roguelikes are single-player games, MUDs are multiplayer games. Roguelikes usually feature permadeath, MUDs rarely do. It's the difference between an RPG and a MMORPG.

Also Nethack offers a turn based top down gameplay simulated with ASCII chars, MUDs are real time or almost real time based with a wall of text and prompt based, like a text adventure.

I don't think anyone plays MUDs with plain telnet. Just use a client with an automapper like tintin++

Or kbtin.

On top of what others have said, the advantage to using one of these remote versions (besides removing the need for a local install), is that they often provide support for arcade-style features like leaderboards and the ability to watch other players games either live or as replays.

In proud to say that I graduated in CS in the 90s despite countless hours of Nethack. I never ascended though.

And I can say I never got fired playing rogue when I started out programming.

Next time, we’ll learn how one of the most popular of all the slick commercial games of the late 1990s grew out of this odd little corner of hackerdom…

I wonder what game he's talking about. Have to circle back sometime or subscribe to their RSS feed.

Diablo, I guess.

It’s got quite clear rogue influences.

at the time when Diablo came out it was clear it was a polished, graphical version of Nethack

It was more like a polished graphical off-shot of Angband. But I wouldn't call it a "version of it" anymore than I'd call Battlefield a version of Doom (or any platformer a version of Super Mario), they are simply games in the same (or similar) genre.

(It’s Diablo.)

[on open source] As for the ideas they introduced into the public discourse: they were real, valid, and in many ways incredibly valuable, but in the end they would be woven into the fabric of existing corporate-software production practices rather than burning down the old ways wholesale.

Ouch

For rigid ideology seldom makes a good fit with the real world; pragmatically mixed national economies, for example, succeed vastly better than dogmatically capitalist or communist ones.

Agree with this too, you need diversity in ideas / ideology

Yeah, I work in an investment bank in Hong Kong, and I agree with both statement: we are more successful than most countries, including China, because as a people we are extremely pragmatic: we have free education to provide banks with an adequate workforce, we have free hospital so that bank employees do not overly concern themselves with the wellbeing of their parents and children, we have freedom of expression because we should not be distracted from the main point, politically: we should only aim at fighting for more capital to come in, not for freedom of speech to reappear. We only break down when a dogmatic partner, be it China or the US, push us around to trend their side, it's very annoying to witness.

I enjoy a lot remixing open source software into our bank's backend. However, nobody ever cared about what I did in the bank to a point I could send it back to an open source community. I think open source cannot solve problems alone, as the article says: it just solves technical problems but never the ones I need to solve for money. Interestingly some of my least productive colleagues, those who never seem to solve any problem, are very very avid open source contributor. It's like they can write an entire unusable framework for calling databases in 6 whole months, but we still have clients who cannot cross a swap in Korea. They don't seem to know what a swap is, why we want to cross them or where Korea is located. Why ?

I've managed to finish Rogue (version IV, DOS) a few times from many hundreds of games over 25+ years. The secret (hopefully this isn't a spoiler) is to know how to use the scroll of scare monster. I find that, when I'm playing, I enter a state of 'flow' that I suspect might be similar to (or maybe the opposite of?) meditation - my feeling of conscious self disappears - the world is everything around the '@' symbol.

A nice read, and it was nice to learn the back story of rogue. One slight correction: PLATO was at the University of Illinois (Urbana-Champaign), not Chicago. I think there were some terminals at UIC, but not U of C.

There was another rogue clone that I remember grabbing from Usenet back in the day, by Tim Stoehr. It was intended to be as direct a clone as possible (nethack was already a thing by thing by then). I implemented hand grenades in my long-lost copy.

https://www.roguebasin.com/index.php?title=Rogue_Clone

Unfortunately, link rot has set in. Only the Sourceforge link still works:

https://rogueclone.sourceforge.net/

It seems to be a DOS/Windows port.

It just so happens I did some work on "rogue clone" a few weeks ago – this will run on most modern systems (AFAIK) and has a few convenience enhancements: https://github.com/arp242/rogue-clone

cf https://www.cs.princeton.edu/~appel/papers/rogomatic.html "ROG-O-MATIC: A Belligerent Expert System"

(a ca. 1984 AI that would play rogue(6) for you)

Brings back fond memories of stumbling across moria on some unix system I was using when I was an intern for the Air Force back in 91 or so. There's something about that style of gameplay that immediately draws you in (that is, if you're the type that "gets" it).

Also, filfre.net is a treasure.

Nethack was promoted to me as a good way to learn how to cursor around in vi (hjkl keys)

I always played the Tourist and managed to get all the way to Level 3 after I learned how to use the +3 Nikon effectively. Everyone knew who the ghost belonged to as well.

> Not only was it not a failure upon its eventual release, but Windows 2000 evolved in 2001 into the consumer-grade Windows XP

esr was talking about server OSes, and he proved to be right in the end. Today Windows server ecosystem survived only in deeply entrenched corporate niches or as a legacy payload. No one in their right mind will not start a new project on MS SQL or IIS. Even .Net server development moved to Linux eventually.

Wasn't there also SuperRogue and UltraRogue or something?

I used to spend endless hours playing Rogue on a VAX/Unix computer. Then I'm sure I spent a lot playing NetHack. I totally didn't realize there's still new versions coming out, this was maybe 30+ years ago ;)

The leaderboard in Rogue was cool (I'm pretty sure there was one) and a bunch of us battled over it.

Very interesting read.

I didn't expect to also end up reading about the Digital Antiquarian's take on Mao's Cultural Revolution, but I simply had to follow his link in the opening paragraphs of this article.

That was a nice read! I would have loved to take a brief dip into the less hardcore rogue-lite scene that sprung up, but that might have doubled the length again.

It's a surprisingly late place for this in Filfre's overall story.

So much of the early Filfre content was Zork focused, and Zork's Coconut of Quendor is such a direct nod to the Amulet of Yendor, that I would have expected this entry very early on in the Digital Antiquarian.

I am interested to see what narrative threads lead him to hold this subject for such a late exposition.

It's the setup for discussing Diablo.

Very very intriguing. Always liked roguelikes and now I might have a nethack addiction.

There is no cure.

Reply on news.ycombinator.com